The long-term goal of this project is to define mechanisms of cancer invasion triggered by microenvironment cues and driven by integrins ?3?1 and ?6?4. These integrins negotiate interactions of cells with laminin-5 (Ln-5), a major component of epithelial basement membranes. Altered expression of both ?6?4 and Ln-5 is associated with cancer progression. Definition of possible causal mechanisms, though, has been challenging because both loss and acquisition of ?6?4/Ln-5 functions were reported to promote tumorigenesis. In this proposal, we plan to assess the validity of an integrated view reconciling these contrasting data. In Aim 1, we will test the hypothesis that invading cancer cells "pave their way" with autocrine-secreted extracellular matrix, particularly Ln-5. A challenge of conventional views of cancer cell migration, this hypothesis is based on the physiological switch from static adhesion to migration that occurs in remodeling or healing epithelia, and is supported by our preliminary experiments in which cancer cell scattering was inhibited by Ln-5 or ?3?1, but not ?6 antibodies. To test this hypothesis, we will determine whether secreted Ln-5 is essential for cell scattering induced by the motogens LPA and HGF, and whether these motogenic stimuli coordinate upregulation and/or secretion of Ln-5. In Aim 2, we will test the hypothesis that a Ln- 5/?6?4/hemidesmosome axis regulates cell-cell adhesion and tumor progression. Engagement of ?6?4 with Ln-5 antagonizes ?3?1-mediated migration and enhances cell-cell adhesion via ErbB2. This suggests negative regulation of tumor progression by ?6?4, vis-[unreadable]-vis cancer poor prognosis associated with ErbB2 gene amplification. Recent in vivo data, from ourselves and others, suggest a tumor suppressive role for secreted Ln-5 heterotrimers. Our hypothesis is based on the premise that these apparently contradictory data reflect the status of cancer cell anchoring to Ln-5 via hemidesmosomes (HD), regulated by ?6?4 in cells switching from migratory to stationary and vice versa. To test this hypothesis, we will investigate the role of the ?6?4/HD axis in cell-cell adhesion and ErbB2 localization in cell lines genetically manipulated to express normal or oncogenic ErbB2, and siRNA knocked-down of HD components. Readout assays include cell-cell adhesion measurements in vitro and xenogeneic tumor models in vivo. In Aim 3, we will test the hypothesis that the ?2DIII domain of Ln-5 promotes tumor progression through EGFR. Expression of monomeric Ln-5 ?2 chain is consistently associated with invasion by many reports and our own preliminary microarray analyses. The ?2DIII proteolytic fragment binds EGFR and our preliminary data indicate it upregulates resistance to apoptosis and tumorigenesis in vivo. We will test the ?2DIII effects on in vitro morphology and survival and in vivo tumor growth and invasion, by siRNA knockdown and rescue with ?2 chain or ?2DIII fragment variants, and will assess its binding to additional ErbB family members. In summary, the 3 aims will test validity and limitations of an integrated view of the opposing effects on tumor progression by secreted Ln-5 heterotrimers (negative) and monomeric ?2 (positive). Understanding these mechanisms of cancer invasion may point to new molecular targets for controlling cancer invasion and metastasis, the major cause of mortality in cancer.